Rotation invariant pattern recognition based on amplitude-modulated phase-only filter

WU Dan; LU Yuan; LING Yong-shun; MA Chao-jie; LI Yu-bo

doi:null

您当前的位置：

首页 >

文章列表页 >

Rotation invariant pattern recognition based on amplitude-modulated phase-only filter

Article | 更新时间：2020-08-12

- Rotation invariant pattern recognition based on amplitude-modulated phase-only filter
- Optics and Precision Engineering Vol. 17, Issue 3, Pages: 506-512(2009)
- 作者机构：
  
  电子工程学院安徽省红外与低温等离子体重点实验室,安徽合肥,230037
- 作者简介：
- 基金信息：
- DOI：
  CLC： O438;TN713
- Received：29 April 2008，
  
  Revised：23 July 2008，
  
  Published Online：25 March 2009，
  
  Published：25 March 2009
- 稿件说明：
移动端阅览
吴丹, 路远, 凌永顺, 马超杰, 李玉波. 应用振幅调制纯相位滤波器实现旋转不变识别[J]. 光学精密工程, 2009,17(3): 506-512

WU Dan, LU Yuan, LING Yong-shun, MA Chao-jie, LI Yu-bo. Rotation invariant pattern recognition based on amplitude-modulated phase-only filter[J]. Editorial Office of Optics and Precision Engineering, 2009,17(3): 506-512
吴丹, 路远, 凌永顺, 马超杰, 李玉波. 应用振幅调制纯相位滤波器实现旋转不变识别[J]. 光学精密工程, 2009,17(3): 506-512 DOI：

WU Dan, LU Yuan, LING Yong-shun, MA Chao-jie, LI Yu-bo. Rotation invariant pattern recognition based on amplitude-modulated phase-only filter[J]. Editorial Office of Optics and Precision Engineering, 2009,17(3): 506-512 DOI：

摘要

提出了一种基于振幅调制纯相位滤波器(AMPOF)的旋转不变识别新算法。该算法充分利用振幅调制纯相位滤波器区分能力强的优点

结合综合鉴别函数方法

拓展了综合鉴别函数(SDF)在畸变不变模式识别领域的应用。利用各幅训练图像傅里叶变换的振幅和相位信息

逐一构建成子滤波器

并将它们加权求和得到合成滤波器;然后采用迭代算法求解出合理的权重系数。仿真实验结果表明:对于实验采用的目标训练图像和非训练图像

改进算法的相关输出峰值大约提高95%~112%;对于只有少量像素的局部训练图像

改进算法的相关输出峰值大约提高74%。这种优化算法使得振幅匹配和相位匹配更加充分

Abstract

In order to get good general performance of optical matched filtering correlation recognition

a new algorithm based on Amplitude-modulated Phase-only Filter(AMPOF) is proposed for rotation invariant pattern recognition. The algorithm extends the application of Synthetic Discriminant Function(SDF) to distortion invariant recognition efficiently by utilizing good discrimination capacity of the AMPOF and employs both amplitude and phase information of Fourier transforms of training images to construct a series of sub filters

and to get a synthetic filter by weighted sum of sub filters. Finally

the algorithm can obtain the weight coefficients by using a iterative algorithm. The simulation results indicate that the correlation peak value of improved algorithm can be enhanced by 95%~112% for training images and non-training images in experiment

and can be enhanced by 74% for one local training image with a few of pixels. The amplitude matching and phase matching will be much sufficient under new algorithm

consequently

the correlation output gain will be enhanced evidently

and the training data can be reduced to a certain extent.

关键词

Keywords

references

王玉荣,徐鹏,王青圃,等. 光电混合联合变换相关器中各元器件结构参数之间的关系[J]. 光学精密工程, 2005,13(3):376-384. WANG Y R, XU P, WANG Q P, et al.. Relationship between parameters in hybrid optical/electronic joint transform correlator[J]. Opt. Precision Eng., 2005,13(3):376-384. (in Chinese)[2] HORNER J L,GIANINO P D. Phase-only matched filtering[J]. Appl. Opt.,1984,23(6):812-816.[3] MU G G,WANG X M,WANG Z Q. Amplitude-compensated matched filtering[J]. Appl. Opt.,1988,27(16):3461-3463.[4] AWWAL A A S,KARIM M A,JAHAN S R. Improved correlation discrimination using an amplitude-modulated phase-only filter[J]. Appl. Opt.,1990,29(2):233-236.[5] BUTT J A,WILKINSON T D.Wilkinson. Binary phase only reference for invariant pattern recognition with the joint transform correlator[J]. SPIE,2006,6234,62340J:1-11.[6] CASASENT D.Unified synthetic discriminant function computational formulation[J]. Appl. Opt.,1984,23(10):1620-1627.[7] JEONG K H,POKHAREL P P,XU J W, et al.. Kernel based synthetic discriminant function for object recognition . Proceedings of ICASSP, 2006:765-768.[8] 沈学举,王永仲,姚广涛,等. 扩展等相关峰综合鉴别函数算法研究[J]. 光学精密工程, 2008,16(1):156-160. SHEN X J, WANG Y ZH, YAO G T, et al.. Study on extended equal correlation peak synthetic discriminant function algorithm[J]. Opt. Precision Eng. , 2008,16(1):156-160. (in Chinese)[9] ALAM M S,AWWAL A A S,WHORTER P M, et al.. Scale invariant amplitude modulated phase-only filtering[J]. Optics&Laser Technology,2000,32:231-234.[10] SABA A M E,ALAM M S,SAKLA W. Enhanced rotation and scale invariant target detection using the fringe-adjusted joint transform correlation[J]. SPIE,2006,6245,62450G:1-12.[11] HORACHE E H,ALAM M S,LOO C H, et al.. SDF-based nonzero-order fringe-adjusted joint transform correlator using binary random phase mask[J]. SPIE,2004,5559:358-368.[12] ngel González-Fraga J, Víctor H.Díaz-Ramírez, Vitaly Kober, et al.. Improving the discrimination capability with an adaptive synthetic discriminant function filter[J]. IbPRIA LNCS,2005,3523:83-90.[13] JARED D A,ENNIS D J. Inclusion of filter modulation in synthetic-discriminant-function construction[J]. Appl. Opt.,1989,28(2):232-239.[14] ENCARNACIN P,IGNASI L,SANTIAGO V, et al.. Generalization of the Jared-Ennis method to complex transmittance objects for the generation of synthetic discriminant function filters[J]. SPIE,2003,5202:227-236.[15] HORACHE E,ALAM M S,LOO C. Optical implementation of SDF-based fringe-adjusted joint transform correlation for invariant pattern recognition[J]. SPIE,2005,5816:152-162.[16] TAO C K. Ill-posed problem of the constant-constraint in optical correlation pattern recognition[J]. J.Opt.Soc.Am. A,1995(12):2114-2121.[17] HUA W SH,XUE R,ZHANG Y, et al.. Experimental investigation of distortion range for synthetic discriminant function filter in optical correlator[J]. SPIE,2005,5642:523-530.[18] MELA C L,IEMMI C.Joint-transform correlator tolerant to distortions of the objects to detect[J]. SPIE,2004,5622:1304-1309.

Views

323

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

No data

Related Author

No data

Related Institution

No data

AI问答

Address：No.3888 Dong Nanhu Road, Changchun, Jilin, China Postal code：130033
Tel：0431-86176855 Email：gxjmgc@ciomp.ac.cn
Technical support is provided by Beijing Founder electronics co., LTD 吉ICP备11002662号-17 京公网安备11010802024621
It is recommended to read the content of this site in Chrome&IE9+. Please switch to extreme mode in browser 360.
Cookies We use cookies to help provide and enhance our service and tailor content. By continuing, you agree to the use of cookies.

⁰